JP3675375B2 - Nonvolatile memory and data rewriting method of nonvolatile memory - Google Patents

Description

[0001]
BACKGROUND OF THE INVENTION
Even if a system failure occurs due to a power failure or the like in a non-volatile memory such as a NAND flash memory provided in an electronic device such as a memory card, the present invention detects a data block in which a defect occurs at the time of restart The present invention relates to a nonvolatile memory capable of returning to a proper storage state and a data rewriting method in the nonvolatile memory.
[0002]
[Prior art]
In recent years, non-volatile memories such as low-cost NAND flash memory suitable for storing large amounts of data have been widely used in electronic devices such as home appliances, portable electronic devices, and memory cards. For example, when a system failure occurs due to a power failure, malfunction, or the like in an electronic device including a flash memory, part of data stored in the flash memory may be destroyed. One method of data protection measures against such a system down is to make the data to be written to the flash memory redundant using, for example, a checksum or CRC (Cyclic Redundancy Check) code, and to flash the system when power is turned on. A checksum is calculated from the data stored in the memory, and it is determined whether or not the data is abnormal.
[0003]
As another method for data protection, an auxiliary non-volatile memory capable of writing at a higher speed than the flash memory is provided in addition to the flash memory, and the latest predetermined number of operating states (status) in the auxiliary non-volatile memory are provided. The bus information related to is stored. FIG. 8 is a block diagram showing a schematic configuration of an electronic apparatus equipped with such a conventional flash memory. In FIG. 8, 101 is an electronic device, 102 is a CPU, 103 is a flash memory provided as a main memory, 104 is an auxiliary non-volatile memory capable of writing faster than the flash memory, and 105 connects the CPU 102 and the flash memory 103. A bus 106 is branched from the bus 105 and connected to the auxiliary nonvolatile memory 104. A signal line 107 supplies a control signal to the auxiliary nonvolatile memory 104.
[0004]
Next, the operation of the electronic device shown in FIG. 8 will be described. For example, in the first half of each operation period of the flash memory 103, the CPU 102 supplies a control signal to the auxiliary nonvolatile memory 104, and the auxiliary nonvolatile memory 104 transmits information on the bus 105 via the bus 106 in response to the control signal. Record. In this bus information recording, only the bus information related to the latest plural statuses is recorded with respect to the status generated by processing such as data writing, reading, and erasing. Therefore, the auxiliary nonvolatile memory 104 does not require a large storage capacity. As a result, in a flash memory that requires a long time for write and erase operations, even if a system failure occurs due to a power failure or the like during operation, bus information related to a predetermined number of statuses immediately before that is recorded in the auxiliary nonvolatile memory. Therefore, by analyzing the bus information recorded in the auxiliary non-volatile memory after system recovery, the operating state at the time of system down is identified, data abnormality in the flash memory is detected, and the flash memory is returned to the normal storage state Let
[0005]
[Problems to be solved by the invention]
A conventional electronic device equipped with a flash memory has a data protection measure function as described above, and the method using the former checksum or CRC code cannot detect the operating state when the system is down. There has been a problem that it is necessary to construct a complicated system in order to perform the process of restoring the system or the process of restoring the data in the flash memory.
[0006]
Further, in the latter method of storing bus information in the auxiliary non-volatile memory, the stored bus information is analyzed to identify the operating state at the time of system down and the data in the flash memory is determined according to the specified operating state. There is a problem that it is necessary to construct a complicated system in order to perform data restoration processing.
[0007]
The present invention has been made to solve the above-described problems, and can easily detect an operation state related to a nonvolatile memory such as a NAND flash memory when the system is down due to a power failure or the like. It is an object of the present invention to obtain a nonvolatile memory capable of surely returning to a normal storage state and a data rewriting method for the nonvolatile memory.
[0008]
[Means for Solving the Problems]
  The present inventionThe physical block address converted by referring to the address conversion table from the logical block addressIn a nonvolatile memory configured with a physical block as a storage unit,Each physical block includes a data area in which data is stored and a redundant area in which various management information is stored.Each of the redundant areas is a logical block information storage area for storing information for specifying a logical block corresponding to the physical block included in the redundant area, and a physical block in which the corresponding logical block is associated with the previous logical block. Each of the stages that occur when performing a data rewrite operation with a physical block included in the redundant area as a data write target, and a pre-used physical block information storage area for storing information specifying the previous used physical block. A status information storage area for storing status information for identifying the operating state;The status information storage area includes first status information indicating that the physical block included in the status information storage area is in an erased state, and data stored in the physical block included in the status information storage area. Second status information indicating a state in which the data of the corresponding previously used physical block has been written but has not been erased, and writing of the data to the physical block included in the status information storage area is completed At least third status information indicating a state in which the data of the corresponding previous used physical block has been erased is stored, and the first status information, the second status information, and the third status information have the same number of bits. Binary represented by one or more bits in a bit string that is represented and represents first status information The second status information is configured by changing the data from “1” to “0”, and binary data related to one or a plurality of bits in the bit string representing the second status information is changed from “1” to “0”. The third status information is configured by changing to "".
[0009]
  Further, the present invention provides a physical block address converted by referring to the address conversion table from the logical block address.It consists of physical blocks as storage units,Each physical block includes a data area in which data is stored and a redundant area in which various management information is stored.Each of the redundant areas is a logical block information storage area for storing information for specifying a logical block corresponding to the physical block included in the redundant area, and a physical block in which the corresponding logical block is associated with the previous logical block. Each of the stages that occur when performing a data rewrite operation with a physical block included in the redundant area as a data write target, and a pre-used physical block information storage area for storing information specifying the previous used physical block. A status information storage area for storing status information for identifying the operating state;The status information storage area includes first status information indicating that the physical block included in the status information storage area is in an erased state, and data stored in the physical block included in the status information storage area. Second status information indicating a state in which the data of the corresponding previously used physical block has been written but has not been erased, and writing of the data to the physical block included in the status information storage area is completed At least third status information indicating a state in which the data of the corresponding previous used physical block has been erased is stored, and the first status information, the second status information, and the third status information have the same number of bits. Binary represented by one or more bits in a bit string that is represented and represents first status information The second status information is configured by changing the data from “1” to “0”, and binary data related to one or a plurality of bits in the bit string representing the second status information is changed from “1” to “0”. The third status information is configured by changing toNon-volatile memoryA data rewriting method, referring to a second step of searching for a physical block in an erased state to identify a physical block to which data is to be written, and an address conversion table associating a logical block and a physical block, A third step of identifying a previous used physical block that is a physical block associated with a logical block to be rewritten, a fourth step of writing data to the physical block to be written, and data of the previous used physical block A fifth step of erasing, and a sixth step of updating the address translation table so that the physical block to be written is associated with the logical block to be rewritten, and the physical block to be written is initially in an erased state. First status information indicating that there is a fourth step is set. Second status information indicating that data writing is completed after the processing is completed is set, and third data indicating that erasure of data of the previous used physical block is completed after the processing of the fifth step is completed. The status information is set, the first status information, the second status information, and the third status information are expressed by the same number of bits, and the binary related to one or a plurality of bits in the bit string expressing the first status information The second status information is configured by changing the data from “1” to “0”, and binary data related to one or a plurality of bits in the bit string expressing the second status information is changed from “1” to “0”. The third status information is configured by changing to.
[0012]
DETAILED DESCRIPTION OF THE INVENTION
Embodiments according to the present invention will be described below with reference to the accompanying drawings. In the following description, in order to clarify the correspondence between each means and each step described in the embodiment of the present invention and each means and each step of the invention described in the claims. Each means and each process of the invention described in the claims corresponding to each means and each process described in the embodiment are indicated by parentheses as appropriate.
[0013]
Embodiment 1 FIG.
FIG. 1 is a diagram illustrating a relationship between a logical block and a physical block related to a storage area of a nonvolatile memory. A typical rewritable nonvolatile memory is a NAND flash memory. This type of flash memory is managed in units of blocks, and data erasure (usually an operation of writing binary data “1” for all bits) is performed in units of blocks. In the present invention, writing is basically performed in units of blocks. However, it can be executed only by converting binary data related to one or more arbitrary bits from “1” to “0”. Data writing can be performed in a smaller storage unit.
[0014]
When accessing the flash memory, as shown in FIG. 1, the given logical block address is converted into a physical block address by referring to an address conversion table stored in a RAM connected to the CPU, for example. The physical block to be erased, written, read, etc. is specified. By adopting a method of accessing by a logical block address in this way, even if a defective physical block occurs, even if the processing operation for the defective block is performed on another free block for replacement, an application program, etc. On the CPU side that executes, it is possible to access using the same logical block address regardless of the presence or absence of block substitution.
[0015]
FIG. 2 is a diagram showing a data structure related to management information stored in a physical block adopted in the present invention. In FIG. 2, 1 is a data area in which normal data is stored, 2 is a redundant area in which data related to management information, error correction codes, and the like related to the physical block is stored. The physical block is composed of 3 is a logical block address storage area (logical block information storage area) for storing the address of the logical block corresponding to the physical block having the redundant area, and 4 is a logical block corresponding to the physical block having the redundant area. Previously used physical block address storage area (previously used physical block information storage area) for storing the address of the physical block previously associated (hereinafter referred to as a previously used physical block). ) 5 is a first error correction code storage for storing error correction codes related to the logical block address stored in the logical block address storage area 3 and the previous use physical block address stored in the previous use physical block address storage area 4 Area 6 is a process executed for a physical block having the redundant area. Status information storage area for storing status information indicating respective operation states of the respective stages that occurs in Zui, 7 is a second error correcting code storage area for storing error correction code according to the data stored in the data area 1.
[0016]
Next, a data rewriting operation executed on the flash memory according to the present invention will be described. First, a schematic procedure of a rewrite operation related to data stored in a predetermined logical block will be described. First, a logical block to be rewritten is specified. Second, search for free physical blocks. Third, rewrite data is written into an empty physical block. Fourth, the data in the physical block that has been associated with the logical block to be rewritten is erased. Fifth, the address conversion table is updated so that the logical block to be rewritten is associated with the written physical block. In this way, when data is rewritten, data is written to a new physical block and the data stored in the previously used physical block is erased. The management information related to the logical block remains in either the physical block to be written or the physical block used before, and the system can be restored without being lost. It is possible to prevent deterioration and extend the life of the flash memory itself.
[0017]
Next, details of the data rewrite operation will be described. FIG. 3 is a flowchart showing the data rewrite operation. FIG. 4 is a diagram illustrating a change in data in a physical block to be erased and a previously used physical block to be erased during data rewriting. Note that the number shown at the left end of FIG. 4 is a status number for identifying the operation state of each stage that occurs in the process of rewriting data. Here, as a precondition, the address of the logical block to be rewritten is L, and the address of the previous physical block that is initially associated with the logical block in the address conversion table is P1. In the following description, a logical block and a physical block accessed by an address n are appropriately referred to as a logical block n and a physical block n, respectively. First, the logical block L to be rewritten is specified (step S1 (first step)). Second, each physical block in the flash memory is searched to detect a free block and specify the address P2 of the free block (step S2 (second step)). In the initial state, “0000” is used as status information in the redundant area of the pre-use physical block P1 to be erased._h"(H is a hexadecimal number), L is a logical block address, and the physical block P is associated with the logical block L one before the physical block P1 is associated with the logical block L as the previous used physical block address. The address P0 of the previous used physical block is stored, and the binary data “1” is stored in the data area and redundant area of the physical block P2, that is, in the erased state (status 0).
[0018]
If the physical block P2 to which data is to be written is specified, the address P1 of the previous used physical block associated with the logical block L is specified by referring to the address conversion table (step S3 (third step )). For the physical block P 2, data is written to the data area 1, address data related to the logical block L is written to the logical block address storage area 3 in the redundant area 2, and the logical block is stored in the previous physical block address storage area 4. The address data related to the previous used physical block P1 to be erased, which was associated with L one time before, is written, and the address data related to the logical block L and the physical block P1 are written in the first error correction code storage area 5. And the error correction code for the data stored in the data area 1 is written in the second error correction code storage area 7 (step S4 (fourth step)). For the status information storage area 6, the status information “FFFF”_h"(First status information) is held. In FIG. 4," 1-> 0 "indicates that data is being written, and the arrow extending in the vertical direction in the figure indicates that the data from the previous status is In addition, regarding the data write operation to the physical block P2, the operation state in the middle of data being written to the physical block P2 is indicated by status 1, and the data write operation to the physical block P2 is performed. An operation state in which data writing has been completed is indicated by status 2 and distinguished from each other.
[0019]
When the data writing to the physical block P2 is completed, the status information “AAAA” is stored in the status information storage area 6 to indicate that the data writing to the previous used physical block P2 is completed._h"(Second status information) is written (step S5). At this time, the same value is held for the data in the other storage areas in the redundant area 2. Also, the status to the status information storage area 6 is stored. As for the information writing operation, an operation state in which data is being written to the status information storage area 6 is indicated by status 3, and status information “AAAAA” is stored in the status information storage area 6._hThe operation state in which "" is completed is indicated by status 4 and distinguished from each other.
[0020]
Status information “AAAAA_hIs completed, the data in the previous used physical block P1 is erased (step S6 (fifth step)). As described above, all the memory cells in the previous used physical block P1 are erased. In the figure, “0-> 1” indicates that data is being erased, and the erase operation of the previously used physical block P1 is as follows. An operation state in the middle of erasing data of the previous use physical block P1 is indicated by status 5, and an operation state in which data erasure of the previous use physical block P1 is completed is indicated by status 6 to be distinguished from each other.
[0021]
When the data erasure of the previous used physical block P1 is completed, the physical block P2 indicates that the erasure of the data stored in the previous used physical block P1 to which the logical block L was associated one time ago is completed. Status information “0000” in the status information storage area 6 in the redundant area 2 of_h"(Third status information) is written (step S7). At this time, the same value is held for the data in the other storage areas in the redundant area 2. Also, the writing to the status information storage area 6 is performed. Regarding the operation, an operation state in the middle of data being written to the status information storage area 6 is indicated by a status 7, and status information “0000” to the status information storage area 6 is indicated._hThe operation state in which "" is completed is indicated by status 8 and distinguished from each other. In the address translation table, the physical block associated with the logical block L is changed from the physical block P1 to the physical block P2. (Step S8 (sixth step)), thereby completing the data rewriting operation.
[0022]
Next, a method for normalizing the data storage state in the flash memory in accordance with the system recovery process performed after a system failure caused by a power failure or operation failure will be described. FIG. 5 is a flowchart showing a storage state normalizing method according to Embodiment 1 of the present invention. When the system is restarted, a search is performed for each physical block that is a storage unit of the flash memory, and first, status information is read from the status information storage area 6 in the redundant area 2 (step S11). Next, the status information is “FFFF”_hIt is determined whether or not the status information is “FFFF” (step S12)._h”Indicates that the data stored in the physical block should be in an erased state. Therefore, the process of erasing the data in the physical block, that is, binary data for all memory cells in the physical block. A process of writing data “1” is performed (step S13), so that even if a system failure occurs due to a power failure or the like in status 1 or status 2 shown in FIG. The written data can be erased and the physical block can be restored to the normal storage state according to the status information.The re-writing of the data that was scheduled to be written to the physical block Host system or network Using a method such as access to the server can be realized by rewriting the correct data.
[0023]
In step S12, the status information is “FFFF”._hIf the status information is not equal to "AAAA"_hIt is determined whether or not the status information is “AAAAA” (step S14)._h”, The data writing to the physical block has been completed, but data erasure of the previous used physical block related to the physical block may not be completed. The data is erased (step S15), and “0000” is stored in the status information storage area 6 of the physical block._h"Is written (step S16). Thus, even if the system down due to a power failure or the like occurs in the status 4, status 5 or status 6 shown in FIG. 4, the status information of the physical block to be searched is changed and the data By erasing the data of the previously used physical block that has not been erased or being erased, the physical block that is the search target and the previously used physical block related to the physical block can be returned to the normal storage state.
[0024]
In step S14, the status information is “AAAA”._hIf the status information is not equal to "0000", the status information is "0000"._h(Step S17). After step S16 is completed, the status information is "0000" in step S17._h”, The logical block address storage area 3 of the physical block to be searched is referenced to identify the corresponding logical block, and the physical block corresponding to the logical block in the address conversion table is the search target. The registered physical block is registered (step S18), thereby reconstructing an address conversion table that has been lost due to a system failure due to a power failure or the like.
[0025]
In step S17, the status information is “0000”._hIf it is not equal to “,” it is determined that the physical block to be searched is an abnormal block, and the address of the physical block is registered in the abnormal block table (step S19). After completing step S18 and after completing step S19, it is determined whether or not the search has been completed for all physical blocks (step S20). Returning to step S11, the same processing is performed for the next physical block, and when the search is completed for all the physical blocks, the processing shown in FIG. Such processing is performed.
[0026]
In the above-described search processing for all physical blocks, for the factor that determines that the physical block is an abnormal block, the status information of the physical block to be written in status 3 is “FFFF”._hTo AAAA_hWhen a system failure such as a power failure occurs while the status is changed to "0000", the status information of the physical block to be erased in status 5 is "0000"_hTo FFFF_hIf a system failure such as a power failure occurs while the status is changed to "AAAAA", the status information of the physical block to be written is displayed in status 7_hTo "0000"_hIt is conceivable that a system failure such as a power failure occurs during the change to “.” Among these factors, the status information of the physical block to be erased in status 5 is “0000”._hTo FFFF_hWhen a system failure such as a power failure occurs during the change to "", it is determined that the block is an abnormal block by performing the process of step S15 shown in FIG. 5 for the corresponding physical block to be written. The physical block is erased, and the status information of the physical block is “FFFF” which is a normal value._hTherefore, it is possible to delete a physical block for which such erasure processing has been completed from the abnormal block table.
[0027]
FIG. 6 is a flowchart showing a search process for erased physical blocks in an abnormal block. First, the physical block address to be searched is specified by extracting the address of the physical block registered from the abnormal block table so that the search is performed for each physical block registered as an abnormal block (step S21). Next, the status information is “FFFF”_hIt is determined whether or not the status information is “FFFF” (step S22)._hIf it is equal to “,” it indicates that the physical block is the previous used physical block for which the erasure process has been completed, so the address related to the physical block to be searched is deleted from the abnormal block table (step S23). And the status information is “FFFF” in step S22._hIf not, it is determined whether or not the search has been completed for all physical blocks registered as abnormal blocks (step S24). If the search has not been completed for all physical blocks, step S24 is performed. Returning to S21, the same processing is performed for the next physical block specified based on the abnormal block table.When the search is completed for all physical blocks, the processing shown in FIG. The following processing related to an abnormal block to be described later is performed.
[0028]
FIG. 7 is a flowchart showing normalization processing related to an abnormal block. First, in order to search for each physical block registered as an abnormal block, the address of the registered physical block is extracted from the abnormal block table corrected by the processing shown in FIG. A block is specified (step S31). Next, the status information is extracted from the redundant area 2 of the physical block to be searched, and the status information and “AAAAA” are extracted._h(Step S32) As described above, data writing in the NAND flash memory is performed by changing binary data “1” to “0”. Change the status information of the target physical block to “FFFF”_hTo AAAA_hWhen the status information has a value different from the original value due to the occurrence of a system failure due to a power failure or the like when the status information is changed to “”, the status information and “AAAA”_hAND with "AAAA"_hIn status 7, the status information of the physical block to be written is changed to “AAAAA”._hTo "0000"_hWhen the status information has a value different from the original value due to the occurrence of a system failure due to a power failure or the like when the status information is changed to “”, the status information and “AAAA”_hAND with "AAAA"_hIs never equal.
[0029]
In view of the above characteristics related to status information, in the next step, status information and "AAAAA"_hAND with "AAAA"_h(Step S33). The logical product is "AAAAA"._hIs equal to “”, data writing to the physical block to be searched has been completed, but data erasure of the previous used physical block related to the physical block has not been completed. The data of the previously used physical block specified by the address stored in the address storage area 4 is erased (step S34) After step S34 is completed and in step S33, the logical product is “AAAA”._hIf it is determined that it is not equal to “”, the status information storage area 6 of the physical block to be searched is set to “0000”._h"Is written (step S35). By executing the processing of step S34 and step S35, even if a system failure occurs due to a power failure or the like in the status 3 shown in FIG. By writing the status information, the physical block to be searched and the previous used physical block can be returned to the normal storage state according to the status information, and the process of step S35 can be performed. Even if a system failure occurs due to a power failure or the like in the status 7 shown in FIG. 4, new status information is written, and the physical block to be searched is brought into a normal storage state according to the status information. Can be restored.
[0030]
Next, the corresponding logical block is specified by referring to the logical block address storage area 3 of the physical block to be searched, and the physical to be searched as the physical block corresponding to the logical block in the address conversion table. A block is registered (step S36). By performing the process of step S36 on all the physical blocks registered as abnormal blocks, it is possible to complete the restoration process of the address translation table that has been lost due to a system failure due to a power failure or the like.
[0031]
Next, it is determined whether or not the search has been completed for all physical blocks registered as abnormal blocks (step S37). If the search has not been completed for all physical blocks, the process returns to step S31 and the same processing is performed on the next physical block specified based on the abnormal block table. If the search is completed for all physical blocks, the process is terminated.
[0032]
As described above, according to the first embodiment, a new physical block that is in an erased state is allocated to a logical block that is a target of data rewriting, and data is written to the physical block. In a flash memory in which data is rewritten by erasing data of a previously used physical block associated with a block, a redundant area 2 includes a logical block address storage area 3, a previously used physical block address storage area 4, Since the status information storage area 6 is configured, even if the system goes down due to a power failure or the like, the physical block that is the target of writing when the system goes down by referring to the status information storage area 6 It is possible to detect the operation status and to record the physical block address before use. Since the previously used physical block related to the physical block can be identified by referring to the area 4, the physical block that may have been destroyed in accordance with the operation state and the previously used physical block related to the physical block Thus, an appropriate restoration process can be performed, and the flash memory can be restored to a normal storage state.
[0033]
The status information includes “FFFF” indicating that the physical block is in the erased state._h"" Indicates that the data in the physical block to be written has been completed, but the data in the corresponding previous used physical block has not been erased._h"0000" indicates a state in which the writing of data to the physical block to be written is completed and the data of the corresponding previous used physical block has been erased._h", Even if data corruption occurs in the physical block, the status information is" FFFF "_h", The physical block to be written may be erased, and the status information is" AAAAA "._h”So that the previously used physical block can be erased, and the status information is changed when the contents of the restoration process required when the system goes down for each status that occurs during the data rewrite operation changes. Therefore, the flash memory can be easily returned to the normal storage state by referring to the status information.
[0034]
Further, as the data rewrite progresses, the status information related to the physical block to be written becomes “FFFF”._h"," AAAA_h"," 0000_hTherefore, even if an abnormality occurs when the status information is changed due to a power failure or the like and the status information does not have a value that should be taken, the status information changes to constitute the status information. In view of the fact that the binary data relating to a predetermined bit of the bit string is changed from “1” to “0”, the status information in which an abnormality has occurred and “AAAAA”_h”And evaluate the logical product to obtain status information“ FFFF ”._hTo AAAA_h"When the system is changed to"_hTo "0000"_hIt is possible to determine whether a system failure has occurred when changing to "", and it is possible to detect the operating state of the flash memory in detail with a simple determination method, and to store the flash memory correctly and reliably. There exists an effect that it can return to a state.
[0035]
Note that the flash memory and the data rewriting method of the flash memory described in the first embodiment are not intended to limit the present invention, but are disclosed for the purpose of illustration. The technical scope of the present invention is defined by the description of the scope of claims, and various design changes can be made within the technical scope described in the scope of claims. For example, status information is “FFFF_h"," AAAA_h"," 0000_hIt is not limited to "", but is changed by changing binary data related to one or more bits in the bit string representing the first status information indicating the erased state from "1" to "0" In the bit string representing the second status information, second status information indicating a state in which data writing to the physical block is completed but the data of the corresponding previous used physical block is not erased is configured. A state in which the data in the physical block to be written is completed by changing the binary data related to one or a plurality of bits from “1” to “0” and the data in the corresponding physical block before use is already erased If the third status information indicating the first status information is configured, an arbitrary bit is used as the first status information, the second status information, and the third status information. It is possible to set up.
[0036]
【The invention's effect】
As described above, according to the present invention, a physical block including a data area in which data is stored and a redundant area in which various management information and the like are stored is configured as a storage unit, and a logical block to be rewritten is a logical block. On the other hand, a new physical block that was in the erased state is allocated and data is written to the physical block, and data is rewritten by erasing the data of the previously used physical block that was previously associated with the logical block. In a nonvolatile memory, each redundant area stores a logical block information storage area for storing information for specifying a logical block corresponding to a physical block included in the redundant area, and information for specifying a corresponding previous use physical block. The physical block information storage area before use and the physical block containing the redundant area Status information storage area that stores status information that identifies the operating status of each stage that occurs when performing the data rewrite operation that is the target of data write. Even if an error occurs, the operation status of the physical block that was the target of writing at the time of system down can be detected by referring to the status information storage area, and the previous use physical block information storage area can be referenced. Since the previously used physical block related to the physical block can be specified, the physical block that may have been corrupted due to the operation state and the appropriate restoration process for the previously used physical block related to the physical block To restore the nonvolatile memory to a normal storage state. There is an effect that it is.
[0037]
According to the present invention, the status information storage area includes the first status information indicating that the physical block included in the status information storage area is in the erased state, and the physical block included in the status information storage area. Second status information indicating a state in which data writing has been completed but the data of the corresponding previous used physical block has not been erased, and data writing to the physical block included in the status information storage area has been completed And at least the third status information indicating the state in which the data of the corresponding previously used physical block has been erased is stored. When the content of the restoration process required in case of downtime changes, the status information Since There are adapted to be changed, an effect that can be easily returned to the non-volatile memory in the normal storage state by referring to the status information.
[0038]
According to the present invention, the first status information, the second status information, and the third status information are expressed by the same number of bits, and binary related to one or a plurality of bits in the bit string expressing the first status information. The second status information is configured by changing the data from “1” to “0”, and binary data related to one or a plurality of bits in the bit string expressing the second status information is changed from “1” to “0”. Since the third status information is configured by changing to, there was an abnormality that the status information did not have the value that should be originally taken due to a system failure due to a power failure or the like when changing the status information However, binary data related to one or more bits in the bit string constituting the status information is changed by the status information. Considering that it is implemented by changing from “1” to “0”, the logical product of the status information in which the abnormality has occurred and the bit string representing the second status information is calculated and the logical product is evaluated. Whether the system down occurred when changing the status information from the first status information to the second status information or when the status information was changed from the second status information to the third status information It is possible to determine whether it has occurred, it is possible to detect in detail the operation state of the nonvolatile memory by a simple determination method, and the nonvolatile memory can be quickly and surely restored to a normal storage state There is an effect.
[Brief description of the drawings]
FIG. 1 is a diagram illustrating a relationship between a logical block and a physical block related to a storage area of a nonvolatile memory.
FIG. 2 is a diagram showing a data structure related to management information stored in a physical block.
FIG. 3 is a flowchart showing a data rewrite operation in a nonvolatile memory.
FIG. 4 is a diagram illustrating a data change in a physical block when data is rewritten.
FIG. 5 is a flowchart showing a storage state normalizing method according to Embodiment 1 of the present invention;
FIG. 6 is a flowchart showing a search process for erased physical blocks in an abnormal block.
FIG. 7 is a flowchart showing normalization processing related to an abnormal block.
FIG. 8 is a block diagram illustrating a schematic configuration of a conventional electronic device including a flash memory.
[Explanation of symbols]
1 data area, 2 redundant area, 3 logical block address storage area (logical block information storage area), 4 previous use physical block address storage area (previous use physical block information storage area), 5 first error correction code storage area, 6 Status information storage area, 7 Second error correction code storage area

Claims (5)

In a nonvolatile memory configured with a physical block of a physical block address converted by referring to an address conversion table from a logical block address as a storage unit,
Each physical block includes a data area in which data is stored and a redundant area in which various management information is stored.
Each of the redundant areas stores a logical block information storage area for storing information specifying a logical block corresponding to a physical block included in the redundant area;
A previous use physical block information storage area for storing information for specifying a previous use physical block which is a physical block to which a corresponding logical block was associated one time before;
A status information storage area for storing status information for identifying an operation state of each stage that occurs when performing a data rewrite operation with a physical block included in the redundant area as a data write target ;
In the status information storage area, the first status information indicating that the physical block included in the status information storage area is in an erased state and the writing of data to the physical block included in the status information storage area are completed. However, the second status information indicating a state in which the data of the corresponding previously used physical block is not erased, and before the data writing to the physical block included in the status information storage area is completed At least third status information indicating a state in which the data of the used physical block has been erased;
The first status information, the second status information, and the third status information are expressed by the same number of bits, and binary data relating to one or a plurality of bits in the bit string expressing the first status information is changed from “1”. The second status information is configured by changing to “0”, and the binary data related to one or a plurality of bits in the bit string representing the second status information is changed from “1” to “0”. 3 status information is configured
A non-volatile memory characterized by that. The physical block of the physical block address converted by referring to the address conversion table from the logical block address is configured as a storage unit, and each physical block stores a data area in which data is stored, various management information, and the like. a that redundant area, each of the redundant area, the logical block information storage area for storing information for identifying the logical block corresponding to the physical block including the corresponding redundant area, the corresponding logical block before one A data rewrite operation is performed with a pre-used physical block information storage area for storing information identifying the previously used physical block that is the associated physical block and a physical block included in the redundant area as a data write target. Status information that stores status information that identifies the operating state of each stage that occurs And a storage area, the above-mentioned status information area, and the first status information indicating that the physical block containing the information concerning the status storage area is in the erased state, a physical block containing the information concerning the status storage area The second status information indicating that the data of the corresponding previous used physical block has not yet been erased, and the writing of data to the physical block included in the status information storage area. Is completed and at least third status information indicating a state in which the data of the corresponding previously used physical block has been erased is stored, and the first status information, the second status information, and the third status information include One or more bits in the bit string expressed by the same number of bits and representing the first status information The second status information is configured by changing the binary data from “1” to “0”, and the binary data related to one or a plurality of bits in the bit string representing the second status information is changed from “1” to “0”. A method of rewriting data in a nonvolatile memory in which the third status information is configured by changing to 0 ″ ,
A second step of searching for a physical block in an erased state to identify a physical block to which data is to be written;
A third step of referring to an address conversion table associating a logical block with a physical block and identifying a previous use physical block that is a physical block associated with the logical block to be rewritten;
A fourth step of writing data to the physical block to be written;
A fifth step of erasing the previously used physical block data;
And a sixth step of updating the address translation table so as to associate the physical block to be written with the logical block to be rewritten,
For the physical block to be written, first status information indicating that it is initially in an erased state is set, and second status information indicating that data writing has been completed after the processing of the fourth step is completed. Third status information indicating that the erasure of the data of the previously used physical block is completed after the completion of the process of the fifth step is set;
The first status information, the second status information, and the third status information are expressed by the same number of bits, and binary data relating to one or a plurality of bits in the bit string expressing the first status information is changed from “1”. The second status information is configured by changing to “0”, and the binary data related to one or a plurality of bits in the bit string representing the second status information is changed from “1” to “0”. 3. A method for rewriting data in a nonvolatile memory, wherein the status information of 3 is configured. Perform a search for each physical block that is a storage unit of the nonvolatile memory, read the status information from the status information storage area in the redundant area,
Based on the read status information, it is determined whether the data stored in the physical block should be in an erased state,
If the read status information indicates that the data stored in the physical block should be in an erased state, erase the data in the physical block,
In addition, when the read status information does not indicate that the data stored in the physical block should be in an erased state, writing of data to the physical block is completed, but the physical block It is determined whether there is a possibility that the data erasure of the previous use physical block related to the previous use physical block has not been completed. In addition to erasing data, write status information indicating that the data of the previous used physical block has been erased to the status information storage area of the physical block, and refer to the logical block address storage area of the physical block being searched The corresponding logical block is identified and the address translation table corresponds to the logical block. Register the physical block that is to be searched for as a block,
When the read status information indicates that data writing to the physical block has been completed but there is no possibility that data erasure of the previous physical block related to the physical block has not been completed. Determining whether or not the read status information indicates a state in which the data of the previously used physical block has been erased in the status information storage area of the physical block;
If the read status information indicates that the data of the previous physical block has been erased in the status information storage area of the physical block, the logical block address storage area of the physical block to be searched 3, the corresponding logical block is identified, and the physical block to be searched is registered as the physical block corresponding to the logical block in the address conversion table.
Further, if the read status information does not indicate that the data of the previous used physical block has been erased in the status information storage area of the physical block, the physical block to be searched is an abnormal block. Determine that there is, and register the address of the physical block in the abnormal block table
The data rewriting method for a nonvolatile memory according to claim 2. In addition, the physical block address to be retrieved is identified by taking out the address of the registered physical block from the abnormal block table,
Determine whether the status information indicates that the previous used physical block has been erased,
When the status information indicates that the used physical block has been erased, the address related to the physical block to be searched is deleted from the abnormal block table.
The data rewriting method for a nonvolatile memory according to claim 3. Furthermore, the physical block address to be searched is specified by taking out the address of the registered physical block from the corrected abnormal block table,
Status information is extracted from the redundant area of the physical block to be searched, and writing of data to the physical block to be searched is completed based on the status information. Determine whether data erasure of the used physical block is in an incomplete state,
When the data erasure of the previous used physical block related to the physical block has been completed, status information indicating that the data of the previous used physical block has been erased is stored in the status information storage area of the physical block being searched. When writing or data erasure of the previous used physical block related to the physical block is not completed, the previous used physical block specified by the address stored in the previous used physical block address storage area Erasing data and writing status information indicating that the data of the previous physical block has been erased to the status information storage area of the physical block being searched
The data rewriting method for a nonvolatile memory according to claim 4.

Images